Date:         Mon, 11 Apr 2005 15:09:47 -0400
Reply-To:     Michael Kruger <aa3657@wayne.edu>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         Michael Kruger <aa3657@wayne.edu>
Subject:      Puzzling analyssi question concerning sperm-donors
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

I've been presented with some data from our director of In-Vitro
fertilization program that has me stumped as far as how to analyze the data.

Essentially, we have n=44 women from whom eggs have been 'harvested' for
IVF (in-vitero fertilization). Two (2) separate sperm donors have been
used for
fertilizing each woman's eggs. The question being asked: Is there a
distinct and measurable 'male effect"? A contribution that can be
quatified from the different sperm donors.

A summary of the design is as follows:

_Females_       _Male 1_         _Male 2_        _Outcome Measure
(Pregnancy or No pregnancy)_
  1                    1                  2                 (0-No
Pregnancy, 1- Pregnancy)
  2                    3                  4
  3                    5                  6
  4                     7                 8
...
....

44                    87                  88


Assuming that the eggs are the same for each of the members of a male
donor pair  (which is not a fair assumption to make), how might one
analyze this data to determine if there is a 'male effect'?
   Thinking about it some form of Sign Test might work. If the outcome
from each pair of males is the same, there is no difference (-) but if
the outcomes are
 different  for a pair of male donors (+) then there is a difference or
a 'male effect'. Any thought or suggestions would be appreciated

--
Michael Kruger
"A True Prince"
Statistical Analyst
C.S. Mott Center
Dept. of OB/GYN
Wayne State University School of Medicine
(313)-577-1794